A public, evidence-backed record of Kongregate rankings and observed game play counts, assembled from archived Kongregate pages, archived per-game metrics.json captures, and current category pages.
Explore the animated ranking chart | Open the companion Google Sheet
This is the completed Kongregate/Wayback v1 dataset. The final audit classifies every remaining recovery priority, verifies that all ranked months have aggregate as-of play-count coverage, and finds no true play-count decreases in the published series.
- 49,982 ranked-list rows from 2007-01-20 through 2026-07-04
- 2,998 canonical games that reached the top 20 in an observed ranking
- 44,906 ranked rows (89.84%) with an observed play count on or before the ranking date
- 9,124 per-game history observations across 2,687 games
- 26,438 observed play-count records used by the chart
The animated chart reads its data directly from this repository at runtime: outputs/kongregate_ranked_games/play_count_bar_chart_race_data.json. The Google Sheet is a companion workbook, not the chart's data source.
The remaining 5,076 ranked rows without an aggregate as-of count are documented rather than silently filled. The final audit places each affected game in an evidence bucket: an exact metrics archive begins after the missing period, no usable archived metrics capture exists for the period, archived pages expose dynamic placeholders without a count, or no usable archived page capture exists.
This is therefore complete for the available Kongregate and Wayback evidence. Filling those final gaps would require a materially different source, such as third-party records, developer archives, or private platform data.
Useful starting points:
data/processed/ Published tables, CSVs, JSON exports, and coverage audits
outputs/ Workbook and static chart artifacts served by GitHub Pages
scripts/ Extractors, recovery probes, builders, and QA checks
logs/ Generated machine-readable and narrative run reports
logs/progress/ Historical checkpoint index and release notes
index.html GitHub Pages entry point
The most useful data files are:
data/processed/ranked_games.csv: ranking date, game, rank, source, and direct listing-count observations.data/processed/ranked_games_observed_plays.csv.gz: ranked rows enriched with the highest observed count on or before each ranking date.data/processed/game_play_history.csv: per-game metrics, live metrics, and archived game-page observations.data/processed/mini_catalog.csv: the canonical top-20 mini catalog and recovered game IDs.data/processed/game_lifecycle_catalog.csv: first/last observed dates, source evidence, and cautious removal evidence.data/processed/game_taxonomy.csv: versioned genre, engagement, social-platform review, developer-group, and series-group labels with confidence and evidence fields.outputs/kongregate_ranked_games/kongregate_ranked_games.xlsx: workbook containing the published tables and reports.
logs/ intentionally keeps generated reports at a stable, flat path so the data pipeline can rewrite them without special handling. The logs guide groups them by purpose for humans, and the progress index keeps the long checkpoint trail out of this README.
The scripts are resumable, and raw Wayback caches are intentionally not committed. A normal refresh uses the existing processed snapshot as its starting point:
python3 scripts/extract_ranked_games.py
python3 scripts/build_mini_catalog.py --top-n 20
python3 scripts/build_ranked_games_observed_plays.py
python3 scripts/scan_data_quality.py --as-of 2026-07-05
python3 scripts/build_game_lifecycle_catalog.py
python3 scripts/build_game_taxonomy.py
python3 scripts/summarize_no_history_evidence.py
python3 scripts/build_v1_completion_audit.py --processed-through-priority-rank 189
node --max-old-space-size=8192 scripts/build_ranked_games_workbook.mjs
node scripts/build_play_count_bar_chart_race.mjsThe recovery tools in scripts/ can continue targeted Wayback work, but the v1 audit should be refreshed after any data change before treating a new snapshot as published.
The scripts and other software are available under the MIT License. Original compiled data, reports, visualizations, and documentation are available under CC BY 4.0. See NOTICE.md for source and attribution guidance; no rights are granted to the underlying games, trademarks, or third-party archive material.